Is 24v and 56v possible with a charge pump?

[CheapPedalCollector]

I'm not sure where to even look on designing charge pump circuits, any pointer would be appreciated. Books, online, datasheet circuit examples.

I want make 24v so I can make my own EP booster that runs at the right voltage, and I want 56v so I can replace the power supplies in my Morley pedals to get rid of the hum and make them safer too.

Also is it possible to make them in parallel with 180 phase shift so the oscillator frequency ripple can be cancelled at the output? I often hear it even in the best power supplies as ring modulation or high pitched squeeling that drives me nuts.

[Rob Strand]

It is possible to do higher voltages with charge pumps but using voltage multiplier schemes.  The two common types are Coc-k-croft-Walton and Dickson.

Keep in mind as you boost the output voltage the input current (and switch currents) must increase by a factor equal to the output to input voltage ratio - essentially preserving power.   Simple multipliers actually have voltage loss due to the diodes so will need to up the voltage ratio to compensate.

Carefully choose the switch frequency to keep whine out the way.

Once you get beyond a certain power you end up scrounging to find driver which can supply enough current.

In the end you can end-up with a simple but messy looking circuit compared to say a boost converting using a dedicated IC, or dedicated IC + single MOSFET.

[Clint Eastwood]

For 24V you can use this circuit:



It features a 555 timer IC, I actually use a TLC555, is a cmos variant. The advantage is that it uses very little current by itself, so your battery will last longer.
The maximum current draw is about 10 mA, beyond that the voltage will start to drop below 24v. With little current draw, you might get away with 1uF (smd?) capacitors in the doubler part of the circuit.
The oscillator frequency I don't remember exactly, but is 30-40khz or so. Well above audio anyway.

[mmessmore]

You may build something that takes 9v, but can't provide enough current to be useful.  The 56v sounds most problematic.  You can increase voltage, but you can't increase the available Wattage.

You really need to know the current draw to make sure your switching supply and charge pump circuit can provide that (with a little headroom).

Just a thought, but you may want to go the other way: down to those voltages.  You can buy "laptop" power supplies for them.  You can find them in 24v and 56v (kinda surprised by the latter).

You can also find a ton of resources on DIY.  Look for "bench supply."  If you're paranoid about noise, you could go non switching, especially if you have an old receiver or something to scavenge a transformer from.  It'd be heavy to gig with, though.

Just some thoughts, though.  I don't know what the current requirements of the circuits are.  It may be easier to go from 12v or 18v if you're meaning 9v as "pedalboard power supply."

[amptramp]

You might be better off using a flyback power supply so that switching the main power transistor on occurs at minimum current, meaning less dissipation in the power transistor.  When the flyback switches off, you get a noise spike and some dissipation, but not as much as from a charge pump and the effects can be filtered out.  And flyback topology can be the optimum from low current to about 50 watts.  Another advantage of a multiple-output flyback is any output winding can be used for regulation because all the output voltages track each other and all outputs can be isolated from the input power, solving any ground loop problems.  Flyback converters can be bought as standalone items.

With a charge pump, maximum current is drawn during switching meaning that even if you have a clever push-pull pair of charge pumps, the noise will be there and the efficiency will be poor.

[FSFX]

FWIW - these are some 555 based Dickson type of circuits I use to provide about 45v to some guided missile boost/overdrive pedals that I built. I normally run these with a 12v input.
Despite all the talk of 'whine and noise', these run at 8.5kHz (below the FCC 9kHz threshold) and are really quiet with no detectable noise from them leaking through to the signal path and nothing measurable higher than 100dB down on the signal. However, that may just be due to the good shielding, filtering and layout in my pedal design.

[Mark Hammer]

Just out of curiosity, what would any of us need 56V for?

[FSFX]

Just out of curiosity, what would any of us need 56V for?

It's a nice voltage to run some tubes (valves) at rather than starved plate designs or 200 volts. It is a good compromise for safety and performance.

It is also a handy voltage to regulate down to 48v for phantom power.

[CheapPedalCollector]

Oh wow, thank you for all the responses!

Mark, old chrome Morley pedals have internal 56ish volt power supplies that are AC powered, have no ground, and use a 1/2 wave voltage doubler. They are very noisy and unsafe, I really love my Power Fuzz Wah and the modern ones just don't sound the same. I also have a Volume Boost and a Power Wah Boost. I really want to make them quieter, and safer to operate. I have converted them with minor mods to 3 prong power cables and added a 500ma fuse to them, but they still hum due to the transformer even with some small filtering upgrades I did. I want to basically make some drop in replacements for them.

I am aware of the current issues, and yes for this particular one I was thinking 18v or even 24v input as this is available in some pedal board power supplies. The Morley one would be the most troublesome not only for voltage but the current draw of the light bulb, I need to check how much current the pedals draw. I did want to avoid having to use any special power supply like a laptop one, and I do not want to make a switching supply which I do know how to make already. They tend to fail catastrophically when they do and destroy everything they are connected to. They also really need active cooling, and that's noisy and impractical.

I figured someone had already done a 24v supply, thanks for that.

I'll look into the flyback and dickson circuits.

Google fails me when trying to find info on actual implementations and theory of construction with various charge pump ics. I really want to know how to design with them, rather than just all the pages that show examples and don't tell you what anything does and no theory of operation out side of the chip being a capacitor switch.

Thank you all again for the responses.

[FSFX]

I really want to know how to design with them, rather than just all the pages that show examples and don't tell you what anything does and no theory of operation out side of the chip being a capacitor switch.

The typical charge pump ICs work in a similar way to how bucket brigade devices (BBDs) work. That is electrical charge is stored on a capacitor and then semiconductor switches disconnect the source of the electrical energy and open another switch to transfer the charge to another capacitor. You probably have already read all about this. They are simple in operation so there is probably not much else that can be said.

As regards designing with them, a lot of people fail to understand that the transfer of the charge at the input ideally requires a capacitor connected very closer to the input terminals of the charge pump to supply a low impedance (and low ESR) source for that initial transfer of charge. Many people have noise issues with charge pumps because they only think about filtering the output rather than having that storage capacitor on the input and filtering the input.

Here is one article that may help you a bit.

https://www.analog.com/en/technical-articles/charge-pumps-shine-in-portable-designs.html 

Also this article talks about reducing noise.

https://www.analog.com/en/technical-articles/simple-methods-reduce-input-ripple-for-all-charge-pumps.html

[Rob Strand]

However, that may just be due to the good shielding, filtering and layout in my pedal design.

It depends on the circuit.  Opamps have good power supply rejection but old school transistors gains stages don't.   That's why some old-school designs need extra filtering when using converters.

Layout and using shielded inductors also helps.

[FSFX]

It depends on the circuit.

OK - disagree with me if you like. I am the one that has built these, used them in various circuits and measured the noise.

If people wish to build poor circuits with low PSRR and inadequate filtering then I have no control over that.

Correctly designed transistor stages can be made to have a good PSRR.

'Old school' designs that were derived from 1960s transistor radio circuits that were originally designed for use with little 9 volt batteries and not 'battery eliminators' or other mains connected power supplies will obviously require some changes to make them compatible. So often you see schematics of circuits posted which have no consideration for any power supply noise or RFI reduction.

[Rob Strand]

OK - disagree with me if you like. I am the one that has built these, used them in various circuits and measured the noise.

I'm not disagreeing.  Different circuits have different PSRR's, that's life, a fact. 

If people wish to build poor circuits with low PSRR and inadequate filtering then I have no control over that.

Me either and there's plenty of posts going over the whole idea of filtering to get rid of noise.

That's why we are here.  To give advise to people who don't know stuff.

Correctly designed transistor stages can be made to have a good PSRR.

People like the sound of the old school designs.  They inherently have poor PSRR.
You often can't change the design itself because that changes the sound, so what you do is wrap
a filter around it.

'Old school' designs that were derived from 1960s transistor radio circuits that were originally designed for use with little 9 volt batteries and not 'battery eliminators' or other mains connected power supplies will obviously require some changes to make them compatible. So often you see schematics of circuits posted which have no consideration for any power supply noise or RFI reduction.

Sure.   Even hum wasn't an issue for pedals until people started using plug-packs/battery eliminators, perhaps late 70's.   Companies like Boss had it all worked out from day one.   For older pedals, people had to learn the hard way.

[ElectricDruid]

Just out of curiosity, what would any of us need 56V for?

Hey, that way we could *reliably* fry op-amps, instead of messing about like newbs!!

[Rob Strand]

Hey, that way we could *reliably* fry op-amps, instead of messing about like newbs!!

Yeah, you could kill a few pedals with that.   Maybe a good idea to use different connectors!

[CheapPedalCollector]

Lets fry some 5v CMOS or NMOS with it instead.

Thank you for the links FSFX. I'll do some reading.

Yes these old things are not designed to be low noise, have proper impedance or any other such modern stuff. They sound good tho, and that's all that matters. A lot of the noise can be mitigated, not all of it, but that has to be accepted as it's just part of the circuit.

I'll think I'll start a different thread with the power supply and the mods I've already done to it, to see if there's anything else I can do to further reduce hum. I'll have to draw it out as I haven't got any transformers in my Eagle library yet.

[amptramp]

FWIW - these are some 555 based Dickson type of circuits I use to provide about 45v to some guided missile boost/overdrive pedals that I built. I normally run these with a 12v input.
Despite all the talk of 'whine and noise', these run at 8.5Hz (below the FCC 9kHz threshold) and are really quiet with no detectable noise from them leaking through to the signal path and nothing measurable higher than 100dB down on the signal. However, that may just be due to the good shielding, filtering and layout in my pedal design.

Add me to the list of people who want to see what you have designed.  I would be quite happy to see an 8.5 KHz power converter that can work in an audio circuit with negligible noise.

[FSFX]

Add me to the list of people who want to see what you have designed.  I would be quite happy to see an 8.5 KHz power converter that can work in an audio circuit with negligible noise.

[duck_arse]

Despite all the talk of 'whine and noise', these run at 8.5Hz (below the FCC 9kHz threshold) .....

Hertz? kiloHertz? 8.5Hz would throb more than it would whine [in my opinion].

[FSFX]

Despite all the talk of 'whine and noise', these run at 8.5Hz (below the FCC 9kHz threshold) .....

Hertz? kiloHertz? 8.5Hz would throb more than it would whine [in my opinion].

OK - thanks for your keen eye pointing that out. I expect that I am  not the first one on this forum to make a typo.
I would expect that you would know it is 8.5kHz. Also my stupid wireless keyboard drops characters now and again.